FACH Congestion

[gunpowder]

Dears,

I want to understand how FACH Congestion (DTCH) affects user experience

we have take some actions but it always have some side effects like:
Increasing FACH BW > Decrease State Transition SR
Enable P2D @ FACH Congestion > Excessive users on R99 State

So i need to know if you face the same behavior and if your networks stress on clearing FACH congestion or not

Thanks :)

[Lazaros]

FACH congestion increases the delay of data transmission, but the biggest problem is that access KPIs like RRC SSR deteriorate. I propose, either to expand with second S-CCPCH or to activate FACH congestion switches (P2D when FACH is congested)

[rfsupp]

Delay of data transmission failures is the most failures that you will notice - Considering DTCH congestion.
Accessibility will be noticed if you congest DCCH.

Increasing FACH BW > Decrease State Transition SR - I dont see the relation why the state transition SR will be affected.
Enable P2D @ FACH Congestion > Excessive users on R99 State - Huawei has another switch that transits the uE directly to HS instead R99 as you mentioned.

Deploy the second SCCPCH and increase the CTFC is the best approach to solve it.

[Muhammad Imran Rafique]

Hi rfsupp,

second SCCPCH is allowed in E//? or which vendors exactly support this.

BR

[electron]

Hi rfsupp,

second SCCPCH is allowed in E//? or which vendors exactly support this.

BR

Hi,

In E/// no we are not able, Huawei allow.

Cheers

[mashi123]

Increasing FACH BW > Decrease State Transition SR : more bandwidth need more power to give full coverage otherwise state transitions may fail due to radio issues

[melkaiy]

guys I want to know if we can use FACH instead of DCH???

thank you for your help.

[melkaiy]

And sirs is the rrc state transition visible in layer 3 messages?

[babak1349]

Hi,

In E/// no we are not able, Huawei allow.

Cheers
hello,
I think in E******* already we have 2nd FACH as default. One will be used as FACH1 for signaling purposes and 2nd one as fACH2 will be used for Ps data tanmission

The point is theoretically there is no limit on FACH/RACH from number of users point of view as RNTI range is huge The problem could be delay on data transmission Besides when the cells are highly loaded there is no other option to push users with bursty data to FACH state in order to have better user preception
Besides too much delay on FACH/RACH PS may end to high FACH drop rate as the users data can not be recovered in NodeB due to huge delay and it might need some hidden parameter modification

FACH/RACH congestion itself doesn't have any meaning except vendors put limitation on load threshold on it ( I think Huawei does it but I am not sure), in E******* as far as I know there is no limitation but there is consequence as I mentioned which appears as high FACH drop

BR

[electron]

hello,
I think in E******* already we have 2nd FACH as default. One will be used as FACH1 for signaling purposes and 2nd one as fACH2 will be used for Ps data tanmission

The point is theoretically there is no limit on FACH/RACH from number of users point of view as RNTI range is huge The problem could be delay on data transmission Besides when the cells are highly loaded there is no other option to push users with bursty data to FACH state in order to have better user preception
Besides too much delay on FACH/RACH PS may end to high FACH drop rate as the users data can not be recovered in NodeB due to huge delay and it might need some hidden parameter modification

FACH/RACH congestion itself doesn't have any meaning except vendors put limitation on load threshold on it ( I think Huawei does it but I am not sure), in E******* as far as I know there is no limitation but there is consequence as I mentioned which appears as high FACH drop

BR

Hi,

Just addition to this comprehensive comment in E******* it is possible to measure FACH1 and total FACH load via statistics by below RBS counters:

pmNoOfTfc1OnFach1

pmNoOfTfc2OnFach1


Cheers

[Weiland86]

Hello electron,

I have been having issues lately with PS drop, all of them related to fach drop, i have been reading several thread about this topic. however i still do not have a clear workflow in order to determine what parameter should be change depending on network symptoms. I have seen you have been making recommendations in most of these treads would you mind explain in more detail the fach drop problem and the different ways to approach it? depending on if we have URA and fast dormancy or if we are using a software release older than W13?

If you need information just let me know what counters and formulas you need. I can provide it.
I am very concern in understand this problem and how to hanlde it.


Thanks in advanced.


Hi,

Just addition to this comprehensive comment in E******* it is possible to measure FACH1 and total FACH load via statistics by below RBS counters:

pmNoOfTfc1OnFach1

pmNoOfTfc2OnFach1


Cheers

[babak1349]

Hello electron,

I have been having issues lately with PS drop, all of them related to fach drop, i have been reading several thread about this topic. however i still do not have a clear workflow in order to determine what parameter should be change depending on network symptoms. I have seen you have been making recommendations in most of these treads would you mind explain in more detail the fach drop problem and the different ways to approach it? depending on if we have URA and fast dormancy or if we are using a software release older than W13?

If you need information just let me know what counters and formulas you need. I can provide it.
I am very concern in understand this problem and how to hanlde it.


Thanks in advanced.

Hello,
Main PS drop is on FACH right ? You need better to draw Fach drop against FACH erlang prbably you will see when FACH erlang is around 40, the FACH drop increases exponentially.
It is better to activate UE Exception in RNC ( I assume you have E******* RNC) to see which Exception codes occurring for FACH/RACH coomon PS
To solve this problem I think you know how to manipulate hidden parameters
I experienced the ame issue before and I could see from the trace (CTR) Drop call with highl UL RLC unrecoverable and it seems it is due to the procedure in RNC side and it doesn't receive the UL RLC on the expected time and issues such kind of failure


Cheers

[Weiland86]

Hello babak,

I am using E******* equipment, I have enabled UEH exceptions trace, however, at least for me i do not have a clear move in order to mitigate the fails. Regarding hidden parameter i do not know about them.
If you do not mind and you have the time can you explain in more detail the topic?

Lately i have been "solving" fach problem just reducing the traffic in the cell with problems basically pushing traffic to other cells (qoffset2sn) or decreasing cpich power in the cell. So, so far i do not have a clue about hot to solve it properly.

So please if you have comments or can help me, explain in very detail every move. Thanks in advanced.



Hello,
Main PS drop is on FACH right ? You need better to draw Fach drop against FACH erlang prbably you will see when FACH erlang is around 40, the FACH drop increases exponentially.
It is better to activate UE Exception in RNC ( I assume you have E******* RNC) to see which Exception codes occurring for FACH/RACH coomon PS
To solve this problem I think you know how to manipulate hidden parameters
I experienced the ame issue before and I could see from the trace (CTR) Drop call with highl UL RLC unrecoverable and it seems it is due to the procedure in RNC side and it doesn't receive the UL RLC on the expected time and issues such kind of failure


Cheers

[melkaiy]

how can i see the fach messages in layer 3?

[electron]

how can i see the fach messages in layer 3?

Hi,

Maybe DPDCH message will show you some stuff related to FACH


Cheers